Evidence review room

Give the meeting a bounded static-evidence agenda.

TraceMap can help managers, reviewers, architects, and engineers keep a review meeting attached to what static dependency evidence is known, what is partial, and what is still missing.

Public claim level: concept. No public conclusion without evidence. This page describes a review-room agenda for static dependency evidence, not runtime behavior, production traffic, endpoint performance, outage cause, release safety, operational safety, automated impact judgment, or complete product coverage.

Meeting question

What can this room say, and what still needs an owner?

Known evidencePoint to a public-safe proof path, rule ID/evidence tier, coverage label, file path, line span, commit SHA, extractor version, and limitation when those fields are available.
Partial evidenceKeep reduced coverage and analysis gaps visible so the room does not turn a narrow static trail into a broad answer.
Missing evidenceName the absent proof, the owner decision gap, and the next human review step instead of filling uncertainty with confidence.
Bounded languageUse impact wording only when a reducer-backed result and public-safe supporting evidence are present.

Known evidence is reducer-backed and public-safe; partial evidence is reduced-coverage and labeled; missing evidence is an explicit gap for human review.

That sentence is the review-room rule. It keeps the meeting from blending evidence tiers, coverage labels, and owner judgment into a single uncheckable claim.

Agenda

Six columns are enough to keep the conversation honest.

claim

Write the exact statement the room wants to make about a static dependency, changed contract, or review surface.

proof path

Link the statement to a public-safe route, generated summary, report family, or source-backed reference the reviewer can follow.

rule ID/evidence tier

Name the deterministic rule and evidence strength so semantic, structural, syntax-only, and unknown evidence are not treated alike.

coverage label

Carry complete, partial, reduced, or unavailable coverage beside the claim instead of burying it in a side note.

limitation

Say what the evidence cannot prove, including runtime behavior, production usage, deployment state, and release approval.

owner decision gap

Assign the remaining human question to source review, tests, telemetry, ownership, release process, or another accountable path.

Room roles

Everyone gets the same evidence, but asks a different question.

ManagersWhich decision is blocked by missing evidence, and which follow-up belongs outside the static packet?
ReviewersWhich rule, tier, supporting ID, file span, and limitation justify this statement?
ArchitectsWhich static paths suggest coupling, and where does the map stop because coverage is reduced?
EngineersWhich source area should be inspected next, and what claim should be avoided until the reducer or owner has proof?

Evidence states

The useful review output is small, named, and limited.

knownA statement has a public-safe proof path, rule ID/evidence tier, coverage label, and stated limitation.
partialA statement has some static support, but reduced coverage or syntax-only evidence changes how strongly it can be used.
missingA statement lacks the proof path needed for the meeting to say more than "needs review."
decisionThe room records who owns the remaining review step instead of asking the static map to approve the change.

Non-claims

The review room is not an operational decision system.

Public-safe material

Bring agenda rows, not private internals.

Safe to discussPublic routes, proof paths, rule/status framing, evidence tiers, coverage labels, limitations, generated summary names, file paths, line spans, commit SHA, and extractor versions when the summary is approved for sharing.
Keep privateSource excerpts, database text, configuration values, credentials, workstation paths, repository remotes, scan output folders, analyzer diagnostics, and private sample identities.
Review firstPrivate repository evidence needs a private scan and human review before a public summary can be written from it.

Incident and review pages answer adjacent questions.

Use incident-call orientation when a live call needs a quick static map around a named surface. Use incident review orientation when the meeting is broader follow-up after the event. The review room is the agenda that keeps either conversation attached to evidence and visible gaps. Use the endpoint review playbook when one endpoint-adjacent packet needs a bounded engineering review decision.